Novel aliphatic N-oxide of naphthalimides as fluorescent markers for hypoxic cells in solid tumor

Article abstract of DOI:10.1016/j.ejmech.2011.04.040

A series of novel aliphatic N-oxide of naphthalimides (A1-A5) were designed and prepared. The N-O group was firstly introduced into the amine side chain tailed to planar naphthalimide chromophore as hypoxic bioreductive marker. Fluorescence image analysis

Full text of DOI:10.1016/j.ejmech.2011.04.040

European Journal of Medicinal Chemistry 46 (2011) 3030e3037
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Novel aliphatic N-oxide of naphthalimides as fluorescent markers for hypoxic cells
in solid tumor
*
*
Hong Yin, Weiping Zhu, Yufang Xu , Min Dai, Xuhong Qian , Yuanli Li, Jianwen Liu
Shanghai Key Laboratory of Chemical Biology, School of Pharmacy, East China University of Science and Technology, Shanghai 200237, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of novel aliphatic N-oxide of naphthalimides (A1eA5) were designed and prepared. The NeO
group was firstly introduced into the amine side chain tailed to planar naphthalimide chromophore as
hypoxic bioreductive marker. Fluorescence image analysis showed that the compounds could be used as
potential markers for hypoxic cells (V79) in vitro especially for A1 with 17 times hypoxic-oxic fluores-
cence differential, which was probably due to the bis-bioreduction mechanism.
Received 31 July 2009
Received in revised form
13 April 2011
Accepted 13 April 2011
Available online 21 April 2011
Ó 2011 Elsevier Masson SAS. All rights reserved.
Keywords:
Naphthalimide
Fluorescent marker
Hypoxic
NeOxide
Bioreduction
1. Introduction
The nitro-compounds (Fig. 1) were suggested for identifying
hypoxic cells by fluorescence [8,9,10,11,12,13,14]. The nitro group
The hypoxia arose in many tumors because of inefficient
delivery of oxygen via the poorly organized tumor vasculature
[1,2,3]. It was clear that tumor hypoxia reduced the efficacy of
radiotherapy and in some cases contributed to local failure of
radiotherapy and tumor regrowth. Hypoxia was also involved in
many common types of normal tissue morbidity [4]. Therefore,
a simple method for measuring tumor hypoxia could diagnose
tumors and facilitate optimal treatment schedules for individual
patients, with the use of modalities designed to overcome or take
advantage of the presence of hypoxic cells in their tumor [5].
A number of methods were proposed for measuring hypoxic cell
fraction in tumors, such as oxygen microelectrodes, histomorpho-
metric analysis and DNA strand breaks. However, most of them
were invasive and not readily available to most investigators [6].
Over the past two decades, one simpler, easier and non-invasive
method for hypoxic cells detection was developed through fluo-
rescent agents, such as 2-nitroimidazole hypoxic markers and
nitroaromatic compounds [7]. Some successful examples were
shown in Fig. 1.
quenched the fluorescence of the aromatic ring system through
energy transfer and electron transfer. But the compound became
more fluorescent on bioreduction of the nitro group to amino group
[12]. Thus the hypoxia-specificity response was resulted from both
the hypoxia-dependence of nitro-group bioreduction and the
increase in fluorescence from bioreduction. Usually, the hypoxic
markers used in model experiments in vitro had large planar
aromatic structures and good intercalating properties [11,12,15],
their fluorescent metabolites localized into the nuclear region of
cells because of their affinity to DNA. The naphthalimide derivatives
were reported as antitumor agents [16] with the strong ability of
binding with DNA, which were first discovered by Braña and
co-workers, and the two famous compounds known as amonafide
and mitonafide had been selected for Phase II clinical trials. Hence,
in this paper, we selected naphthalimide as the DNA binding part
for designing novel markers in solid tumors.
In hypoxic cells, the tertiary amine N-oxides could release the
active cytotoxins primarily by the CYP3A isozyme of NADPH:
cytochrome C (P-450) reductase was inhibited by oxygen, probably
because of direct competition between oxygen and the drug at the
enzyme site [17]. After the N-oxides being bioreduced to the cor-
responding tertiary aliphatic amines, they could bind with DNA and
interfere with topoisomerase function [18,19,20]. But the most
interesting thing is that the fluorescence of the N-oxides could
* Corresponding authors. Tel.: þ86 21 64253589; fax: þ86 21 64252603.
E-mail addresses: yfxu@ecust.edu.cn (Y. Xu), xhqian@ecust.edu.cn (X. Qian).
0223-5234/$ e see front matter Ó 2011 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2011.04.040

H. Yin et al. / European Journal of Medicinal Chemistry 46 (2011) 3030e3037
3031
Fig. 1. Structures of some reported markers for hypoxic cells in solid tumors.
Scheme 2. (a) H2O2 (30%), CH3OH, reflux, 1e3 h.
change much after bioreduction (generally for PET or ICT mecha-
nism), and the reduced amines could ensure good uptake into cells
and tight binding with DNA.
With these in mind, we presented a series of novel markers for
hypoxic cells in solid tumors from tertiary amine N-oxide of
Fig. 2. Novel compounds for potential hypoxic fluorescent markers (A1eA5).
-
+
O
N
N
N
O
O
O
N
O
a
Scheme 3. (a) H2O2 (30%), CH3OH, reflux, 5 h.
R
naphthalimides (Fig. 2), such as amonafide, mitonafide and other
antitumor agents. Herein, NeO group was firstly introduced to the
amine side chain tailed to planar naphthalimide chromophore to
identify hypoxic cells. For compound A1, with two active sites for
bioreduction, we expect that it should show higher selectivity than
A2 and other compounds. Meanwhile, the properties of these
compounds as fluorescent markers were also evaluated.
R
B1 (Mitonafide)
B2 (Amonafide)
A1, R=NO2
A2, R=NH2
Scheme 1. (a) H2O2 (30%), CH2Cl2, reflux, 1 h.
Scheme 4. (i) o-nitrophenol, NaOH, DMF, Cu, reflux 1 h; (ii) Fe, acetic acid, reflux 1 h; (iii) hydrochloric acid, acetic acid, NaNO2, 0e5 ꢀC, CuSO4, HOAc, H2O; (iv) N,N-dimethy-
lethylenediamine, ethanol, reflux 2e3 h; (v) H2O2 30%, CH3OH, reflux, 5 h. 0.64 g (86%).

3032
H. Yin et al. / European Journal of Medicinal Chemistry 46 (2011) 3030e3037
Table 1
a
Spectral data of A1eA5 and corresponding amino compounds.a,b
25
20
15
10
5
Compounds
UV lmax (lg
e
)
FL lmax (F)
A1
A2
A3
A4
A5
B1
B2
1
331 (4.04)
343 (3.53)
379 (4.14)
461 (4.40)
425 (4.38)
316 (3.93)
342 (3.98)
378 (4.0)
368 (<0.001)
548 (0.94)
461 (0.90)
522 (0.17)
473 (0.143)
368 (<0.001)
547 (0.03)
456 (0.26)
510 (0.09)
472 (0.03)
2
6
460 (4.34)
424 (4.18)
a
In absolute ethanol.
b
With fluorescence in sodium hydrate solution as quantum yield standard
¼ 0.97).
(
lex/nm ¼ 366 nm,
4
0
2. Results and discussion
500
550
600
2.1. Chemistry
b
wavelength(nm)
30
25
20
15
10
5
The target compounds were synthesized by using 1,8-naphthalic
anhydride and 4-bromo-1,8-naphthalic anhydride as starting mate-
rial. The synthetic routes were shown in Scheme 1e4, respectively.
a
50
A4
B4
40
30
20
10
0
0
500
550
600
650
wavelength(nm)
Fig. 4. Fluorescence spectra of compound A4 (a) and the corresponding amine 2
(10
ꢂ
10ꢁ6 M); (b) with addition of ctDNA in increasing concentration
(0e400 ꢂ 10ꢁ6 M) in 20 mM TriseHCl (pH 7.5) at room temperature, the excitation
wavelength was 460 nm.
450
500
550
600
650
wavelength(nm)
The following listed the synthetic steps for compound A5 as an
example: 4-bromo-1,8-naphthalic anhydride and o-nitrophenol
were dissolved in DMFand stirred for 1 h under refluxwith NaOH and
Cu as catalysts to give a yellow solid 3. The solid 3 was treated with Fe
powder in glacial acetic acid and refluxed for 1 h to afford khaki solid
4. Then 4 was added into the hydrochloric acid and sodium nitrite at
0 w 5 ꢀC for 1 h, followed by the addition of CuSO4 solution and
refluxed for 0.5 h to give a yellow solid benzo(k,l)xanthene-3,4-
dicarboxylic anhydride 5, which was mixed with N,N-dimethyle-
thylenediamine in ethanol and refluxed for 2e3 h to give the
important intermediate product 6. Finally, the NO group was intro-
duced by oxidation with H2O2 (30%) in CH2Cl2 or methanol under
reflux for 1 w5 h, removalof the solvent gave the desired compounds
A5 in high yields. All of the structures were confirmed by IR, 1H NMR
and HRMS. A strong NeO stretching vibration appeared in the range
of 2349e2338 cmꢁ1 in their infrared spectra, which was the char-
acteristic absorption band of the N-oxide.
b
20
15
10
5
A4
B4
0
450
500
550
600
650
The UVeVis and fluorescent data for these compounds were
measured and shown in Table 1. It was found that the emission
wavelength of the N-oxides was red-shifted about 4 w 5 nm
respectively than the corresponding naphthalimides in absolute
ethanol. It was probably due to the increase of electronic pushing-
pulling effects caused by the introduction of the weak electron-
withdrawing NO group to the side chains.
wavelength(nm)
Fig. 3. Fluorescence spectra in different solvents. The concentration of compound A4
was 10ꢁ4 mol/L: (a) in absolute ethanol; (b) in water (20 mM TriseHCl, pH 7.5). The
black represented A4, and the red represented the corresponding amine 2 (For inter-
pretation of the references to colour in this figure legend, the reader is referred to the
web version of this article).

H. Yin et al. / European Journal of Medicinal Chemistry 46 (2011) 3030e3037
3033
a 100
hypoxic
oxic
b
50
40
30
20
10
0
hypoxic
oxic
80
60
40
20
0
0
1
2
3
4
5
6
0
1
2
3
4
5
6
time (h)
time (h)
hypoxic
oxic
70
d 50
c
hypoxic
oxic
60
50
40
30
20
10
0
40
30
20
10
0
-10
0
1
2
3
4
5
6
0
1
2
3
4
5
6
time (h)
Time (h)
e 70
hypoxic
oxic
60
50
40
30
20
10
0
-10
0
1
2
3
4
5
6
time (h)
Fig. 5. The time courses of accumulation of fluorescent metabolites in V79 Chinese hamster cells incubated with 10ꢁ4 M compounds at 37 ꢀC; (a) A1 (b) A2(c) A3 (d) A4 (e) A5.
Meanwhile, as shown in Table 1, the quantum yields of the
N-oxides were ranged from 0.143 to 0.94 ( ), and that of the cor-
responding amines varied from 0.03 to 0.26 ( ) except A1, which
and it resulted in a high quantum yield than the corresponding
amines. It provided the possibility for our design strategy as fluo-
rescence markers.
F
F
indicated that the fluorescent intensities changed much between
the N-oxides and their corresponding amines. In detail, the
quantum yield of A2 was about 30 folds stronger than its former
compound B2 (amonafide), and A3, A4 and A5 were about 3.5, 1.9,
8.2 folds stronger than their corresponding amines respectively.
The reason was probably that the former amines with nitrogen in
the tertiary amine side chains generated Photo-induced Electron
Transfer (PET) between the planar aromatic ring and the side chain.
When the nitrogen atom on the side chain was oxidized, the
property of the group changed from electron donating to electron
withdrawing, which probably inhibited PET effect to some extent,
In addition, the fluorescence of the target compounds (repre-
sentative curves for A4) in ethanol and water was evaluated
respectively as shown in Fig. 3.
In Fig. 3(a), it was found that the fluorescence intensity of A4 in
absolute ethanol was stronger than that of the related amino
compound 2, but it was converse in water (Fig. 3b). The reasons
were probably associated with several aspects as follows: firstly,
the PET effect of the amino compound 2 was stronger than A4 to
quench the fluorescence in ethanol because of the weak protonic
ability; Secondly, the aliphatic N-oxide A4 with more charges was
easier to accumulate or aggregate than its corresponding amine 2 in

3034
H. Yin et al. / European Journal of Medicinal Chemistry 46 (2011) 3030e3037
Fig. 6. Fluorescence microphotographs of V79 cells incubated with 10ꢁ4 M of A1 at 37 ꢀC. After 4.5 h incubation, scanning was taken. Magnification was 1000ꢂ. (a) scanning was
taken on brightfield cells on oxic condition (incubated in air and 5% CO2); (b) excited at 410 nm, cells on oxic condition; (c) scanning was taken on brightfield, cells on hypoxic
condition (incubated in nitrogen and 5% CO2); (d) excited at 410 nm, cells on hypoxic condition.
Fig. 7. Fluorescence microphotographs of V79 cells incubated with 10ꢁ4 M of A2 at 37 ꢀC. After 4.5 h incubation, scanning was taken. Magnification was 1000ꢂ. (a) scanning was
taken on brightfield cells on oxic condition (incubated in air and 5% CO2); (b) excited at 410 nm, cells on oxic condition; (c) scanning was taken on brightfield, cells on hypoxic
condition (incubated in nitrogen and 5% CO2); (d) excited at 410 nm, cells on hypoxic condition.

H. Yin et al. / European Journal of Medicinal Chemistry 46 (2011) 3030e3037
3035
Fig. 8. Fluorescence microphotographs of V79 cells incubated with 10ꢁ4 M of A4 at 37 ꢀC. After 4.5 h incubation, scanning was taken. Magnification was 1000ꢂ. (a) scanning was
taken on brightfield cells on oxic condition (incubated in air and 5% CO2); (b) excited at 410 nm, cells on oxic condition; (c) scanning was taken on brightfield, cells on hypoxic
condition (incubated in nitrogen and 5% CO2); (d) excited at 410 nm, cells on hypoxic condition.
water, which resulted in the fluorescence quenched much of
compound A4 [21]; Thirdly, different hydrogen bonding with water
would affect their fluorescence intensity: the stronger of the ability,
the weaker of the fluorescence.
higher than that of other compounds. The reason was probably
due to that A1 was a “bis-bioreductive” [24] with two different
independent oxygen-sensitive redox centers: the nitro group and
the NO group. Under the oxygen-inhabitable reduction, the NO
group was reduced by CYP3A isozyme of NADPH: cytochrome C
(P-450) reductase. In hypoxic cells, the fluorescence quenching
group (NO2) was reduced to amine by nitroreductase enzymes in
2-electron steps [25], and the chromophore changed from pull-
ingepulling system to pulling-pushing one, which resulted in
fluorescence enhancement markedly. Therefore, it could be
deduced that the bioreduction of the nitro-group contributed to
the higher fluorescence differential of A1. Compared with A1, the
bioreduction of the NeO group resulted in less hypoxic-oxic
fluorescence differential for A2 and A4. The hypoxic-oxic fluores-
cence differential incubated with A3 and A5 was evaluated as 5.2
and 1.5 times, respectively.
2.2. DNA binding property
The binding properties between the compounds and ctDNA
were evaluated (Fig. 4). The fluorescence data were analyzed to give
the binding constant of A4 and its corresponding amine 2 about
6.62 ꢂ 103 Mꢁ1 and 2.9 ꢂ 105 Mꢁ1 respectively [22]. It was about 15
folds lower of the N-oxide than its former, which might be caused
by the disappearance of hydrogen bond between nitrogen and DNA.
It implied that weaker ctDNA binding probably endowed the
N-oxides lower toxicity and extravascular drug transport properties
[20].
By fluorescence microscopy, we also got the fluorescence
microphotographs of V79 cells incubated with 10ꢁ4 M of A1, A2
and A4, respectively (Figs. 6e8). Because the fluorescence inten-
sities of the bioreductive products of compounds A4 and A5 are
not strong, the high concentrations of the fluorescent probes
(10ꢁ4 M) were used in the experiments. Figs. 6e8 showed that the
microphotographs of A1 showed higher differential fluorescence
between hypoxic and oxic V79 cells than other N-oxides did,
which was in accordance with the differential evaluated with
fluorescence scan ascent above. It was also found that the fluo-
rescence intensities in hypoxic cell cultures were stronger than
that in oxic cells. The results showed that the metabolized prod-
ucts (naphthalimides) in hypoxic culture medium exhibited
stronger fluorescence intensities than that in oxic cell cultures,
where the N-oxides could not be metabolized, which was in
agreement with the fluorescence change in water (Fig. 3).
2.3. Evaluation as fluorescence marker
For the evaluation of these compounds as fluorescence markers,
V79 cell which was more sensitive to fluorescence [23] was
selected. Samples from hypoxic and oxic cell suspensions incubated
with the solution of compounds were taken, and the study of the
time courses of accumulation of fluorescent metabolites in V79
cells incubated with 10ꢁ4 M of A1eA5 at 37 ꢀC was carried out with
fluorescence scan ascent. The results were shown in Fig. 5.
In Fig. 5, it was found that obvious fluorescence differential of
cells could be seen under hypoxic and oxic conditions. After about
4.5 h, the hypoxic-oxic fluorescence differential reached the
maximal value. The hypoxic-oxic ratio of A1eA5 in the fluores-
cence images was about 17, 8.4, 5.2, 1.75, 1.5 times respectively.
The largest hypoxic-oxic ratio of A1 in V79 cells (17 times) was

3036
H. Yin et al. / European Journal of Medicinal Chemistry 46 (2011) 3030e3037
3. Conclusion
1-H), 8.21 (d, J ¼ 7.42 Hz, 1H, 11-H) 8.35 (d, J ¼ 7.31 Hz,1H, 3-H) 8.76
(s,1H, 7-H); IR (KBr, cmꢁ1): 3326, 2365, 2338,1696,1660,1334, 734;
In summary, we described the synthesis and DNA-binding
affinities of the novel aliphatic N-oxide of naphthalimides, and
investigated their abilities as fluorescent markers for hypoxic cells
in solid tumors. From the fluorescence images, the target
compounds were shown to be good markers for hypoxia cells (V79)
in vitro especially for compound A1, which was probably due to bis-
bioreduction with 17 times hypoxic-oxic fluorescence differential.
The other N-oxides with differential of 8.4, 5.2, 1.75 and 1.5 times
were also promising candidate markers for hypoxic cells.
HRMS: C22H18O3N2: (M þ H)þ calculated 391.1116, found 391.1119.
4.1.4. N-Oxide Benzo[k,l]thioxanthene-3,4-naphthalimide (A4)
Compound 2 was synthesized as previously reported [27]. H2O2
(30%, 0.2 mL, 20 mmol) was added dropwise to the naphthalimide 2
(0.75 g, 20 mmol) in methanol (30 mL), and the mixture was
refluxed for 5 h, with viscousness on the bottle. After the solvent
was removed, ethyl acetate was added. And the viscousness solid
was filtered and separated on silica gel (CHCl3/methanol ¼ 10:2) to
give black red solid 0.61 g (78%). m.p. 159e160 ꢀC 1H NMR (CD3OD,
ppm) dH: 3.32 (s, 6H, NOCH3), 3.66 (t, J ¼ 7.20 Hz, 2H, NOCH2), 4.65
(t, J ¼ 7.04 Hz, 2H, CONCH2), 7.45 (s, 3H, 9-H, 10-H, 11-H), 7.58 (d,
J ¼ 8.03 Hz, 1H, 8-H) 8.34 (d, J ¼ 8.08 Hz, 3H, 1-H, 2-H, 7-H), 8.54 (d,
J ¼ 8.20 Hz, 1H, 6-H); IR (KBr, cmꢁ1): 3350, 2361, 2338, 1688, 1649,
1583, 1369, 758.1; HRMS: C22H18O3N2 (M þ H)þ calculated 391.1116,
found 391.1114.
4. Experimental protocols
All the solvents were of analytic grade. 1H NMR was mea-
sured on a Bruker AV-500 spectrometer with chemical shifts
reported as parts per million (in acetone-d6/DMSO-d6/CDCl3,
TMS as an internal standard). Mass spectra were measured on
an HP 1100 LC-MS spectrometer. Melting points were deter-
mined by an X-6 micro-melting point apparatus and uncor-
rected. Absorption spectra were determined on PGENERAL
TU-1901 UVeVIS Spectrophotometer.
4.1.5. 4-(2-Nitrophenoxy)-1,8-naphthalic anhydride (3)
A mixture of 4-bromonaphthalic anhydride (1.02 g, 4.4 mmol),
o-nitrophenol (0.34 g, 8.5 mmol), sodium hydroxide (0.025 g) and
copper powder (0.04 g) was refluxed in DMF (45 mL) for 1 h.
Hydrochloric acid (7.5 mL, 20%) was added to the solution and the
solid was precipitated, filtered and recrystallized in AcOH to afford
the title compound in 82% yield. m.p. 268e269 ꢀC, and m.p.
266e268 ꢀC in literature [28].
4.1. Synthesis of naphthalimide derivatives
4.1.1. N-Oxide 3-nitro-1,8-naphthalimide (A1)
H2O2 (30%, 3 mL, 30 mmol) was added drop-wise to the 3eNO2-
naphthalimide (0.94 g, 3 mmol) in methylene chloride (30 mL), and
the mixture was refluxing for 3 h. After removing most of the
solvent, ethyl acetate was added. The yellow solid was precipitated,
filtered and separated on silica gel (CHCl3/methanol ¼ 10:2), which
yielded pale yellow solid 0.79 g (80%). m.p.184e185 ꢀC .1H NMR
(CD3OD, ppm) dH: 3.33 (s, 6H, NOCH3), 3.70 (t, J ¼ 7.00 Hz, 2H,
NOCH2), 4.69 (t, J ¼ 6.85 Hz, 2H, CONCH2), 8.00 (t, J ¼ 7.65 Hz, 1H, 2-
H), 8.63 (d, J ¼ 8.26 Hz, 1H, 1-H), 8.77 (d, J ¼ 7.24 Hz, 1H, 3-H), 9.19
(d, J ¼ 2.14 Hz,1H, 9-H), 9.33 (d, J ¼ 2.10 Hz,1H, 7-H); IR (KBr, cmꢁ1):
3069, 2353, 2338, 1664, 1341, 800; HRMS: C21H16O5N3 (M þ H)þ
calculated 330.1130, found 330.1103.
4.1.6. 4-(2-Aminophenoxy)-1,8-naphthalic anhydride (4)
A mixture of compound 3 (0.25 g, 0.8 mmol) and iron powder
(0.12 g, 2 mmol) was refluxed in glacial acetic acid (10 mL) for 1 h.
To this brown solution, water (30 mL) was added. The precipitated
yellow solid was filtered and washed with water to afford the title
compound in 94% yield. m.p. 172e174 ꢀC, and m.p.171e172 ꢀC in
literature [28].
4.1.7. Benzo[k,l]xanthene-3,4-dicarboxylic anhydride (5)
A solution of compound 4 (0.5 g, 1.7 mmol) in glacial acetic acid
(12 mL) was treated with hydrochloric acid (1 mL) and sodium
nitrite (1.14 g, 16 mmol in 4 mL water) at 0 ꢀC. After 60 min,
a solution of copper sulfate (1.12 g, 7 mmol) in water (20 mL) was
added. The mixture was refluxed for another 0.5 h and then allowed
to cool. The precipitated solid was filtered, washed with water and
crystallized in DMF to afford the title compound, yield 93.2%. m.p.
156e159 ꢀC, and m.p. 155e160 ꢀC in literature [28].
4.1.2. N-Oxide 3-amino-1,8-naphthalimide (A2)
3eNH2-naphthalimide (0.85 g, 26 mmol) was dissolved in
methylene chloride (25 mL), and H2O2 (30%, 2.7 mL, 26 mmol) was
added dropwise. The mixture was stirred and refluxed for 3 h. After
the solvent was removed, the ethyl acetate was added to give the
orange solid. The solid was filtered and separated on silica gel
(CHCl3/methanol ¼ 10:1), which gave the pure product 0.76 g
(85%). m.p. 160e161 ꢀC 1H NMR (CD3OD, ppm) dH: 3.30 (s, 6H,
NOCH3), 3.64 (t, J ¼ 7.13 Hz, 2H, NOCH2), 4.63 (t, J ¼ 6.96 Hz, 2H,
CONCH2), 7.35 (d, J ¼ 2.38 Hz, 1H, 9-H), 7.57 w 7.61 (m, 1H, 1-H),
7.98 (t, J ¼ 4.51 Hz, 1H, 2-H), 8.03 w 8.04 (m, 1H, 7-H), 8.18 w 8.20
(m, 1H, 3eH); IR (KBr, cmꢁ1): 3416, 3338, 2353, 2326, 1649, 1622,
789; HRMS: C16H18O3N3 (M þ H)þ calculated 300.1348, found
300.1333.
4.1.8. Benzo[k,l]xanthene-3,4-naphthalimide (6)
Compound 5 was dissolved in 20 mL absolute ethanol, with N,
N-dimethylethylenediamine added, the mixture was stirred and
refluxed for 2e3 h. Then the solution was evaporated in vacuum,
and the residue was purified on silica gel chromatography, eluting
with (CHCl3/ethanol, 10:1) to yield the 6, 0.314 g (85%). m.p.
191e192 ꢀC; 1H NMR (CDCl3, ppm) dH: 2.43e2.46 (m, 6H, N(CH3)2),
2.75 (s, 2H, CH2N(CH3)2), 4.38 (t, J ¼ 7.06 Hz, 2H, CONCH2), 7.29 (d,
J ¼ 8.30 Hz, 1H, 11-H), 7.35 (d, J ¼ 7.37 Hz, 1H, 8-H), 7.36 (t,
J ¼ 7.92 Hz, 1H, 9-H), 7.54 (t, J ¼ 7.17 Hz, 1H, 10-H), 7.95 (d,
J ¼ 7.89 Hz, 1H, 7-H), 8.08 (d, J ¼ 7.90 Hz, 1H, 1-H), 8.59 (d,
J ¼ 8.29 Hz, 1H, 6-H), 8.64 (d, J ¼ 7.88 Hz, 1H, 2-H); IR (KBr, cmꢁ1):
2937，2770，1645，1594，1380; HRMS: C22H18N2O3 calculated
358.1317, found 358.1321.
4.1.3. N-Oxide Benzo[b]xanthene [2,1-c]naphthalimide (A3)
Compound 1 was synthesized as previously reported [26]. H2O2
(30%, 0.2 mL, 20 mmol) was added dropwise to the solution of
compound 1 (0.75 g, 2 mmol) in methanol (30 mL). The mixture
was refluxed for 5 h, with viscousness on the bottle. After the
solvent was removed, ethyl acetate was added. The solid was
filtered and separated on silica gel (CHCl3/ethanol ¼ 10:2) to give
pale yellow solid 0.7 g (90%). m.p. 152e153 ꢀC 1H NMR (CD3OD,
ppm) dH: 3.35 (s, 6H, NOCH3), 3.68 (t, J ¼ 7.27 Hz, 2H, NOCH2), 4.58
(t, J ¼ 7.04 Hz, 2H, CONCH2), 7.50 w 7.54 (m, 2H, 9-H, 10-H), 7.66
(t, J ¼ 7.74 Hz,1H, 2-H), 7.87 w 7.89 (m, 1H, 8-H), 8.10 w 8.12 (m, 1H,
4.1.9. N-Oxide Benzo[k,l]xanthene-3,4-naphthalimide (A5)
H2O2 (30%, 0.2 mL, 20 mmol) was added dropwise to 6 (0.74 g,
2 mmol) in methanol (30 mL), the mixture was refluxing for 5 h with
viscousness on the bottle. After the solvent was removed, ethyl

H. Yin et al. / European Journal of Medicinal Chemistry 46 (2011) 3030e3037
3037
[6] (a) A.W. Fyles, M. Miolsevic, R. Wong, M.C. Kavanagh, M. Pintilie, A. Sun,
W. Chapman, W. Levin, L. Manchul, T.J. Keane, R.P. Hill, Radiother. Oncol. 48
(1998) 149e156;
acetate was added. The viscousness solid was filtered and separated
on silica gel (CHCl3/methanol ¼ 10:1) to give pale yellow solid 0.64 g
(86%). m.p. 169e170 ꢀC; 1H NMR (CD3OD, ppm) dH: 3.35 (s, 6H,
NOCH3), 3.69 (t, J ¼ 7.07Hz, 2H, NOCH2), 4.63 (t, J ¼ 6.97Hz, 2H,
CONCH2), 7.23 (d, J ¼ 8.33 Hz, 1H, 7H), 7.32 (d, J ¼ 8.25 Hz, 1H, 8-H),
7.35 (t, J ¼ 8.19 Hz, 1H, 9-H), 7.55 (t, J ¼ 8.49 Hz, 1H, 10-H), 7.97 (d,
J ¼ 7.95 Hz,1H,11-H), 8.12 (d, J ¼ 8.02 Hz,1H,1-H), 8.43 w 8.46(m, 2H,
2-H, 6-H); IR (KBr, cmꢁ1): 3536, 2349, 2338, 1645, 1590, 1380, 777;
HRMS: C22H18O4N2 (M þ H)þ calculated 391.1345, found 375.1347.
(b) H.A. Al-Hallaq, J.N. River, M. Zamora, H. Oikawa, G.S. Karczmar, Int. J.
Radiat. Oncol. Biol. Phys. 41 (1998) 151e159;
(c) A.Y. Karen, B.P. Siham, M.L. Rachelle, Q.B. Darrell, C.F. Matthew, S.B. James,
E.H. Gerald, J.D. Chapman, Int. J. Radiat. Oncol. Biol. Phys. 33 (1995) 111e118;
(d) A.S.E. Ljungkvist, J. Bussink, P.F.J.W. Rijken, J.A. Raleigh, J. Denekamp,
A.J. van der Kogel, Int. J. Radiat. Oncol. Biol. Phys. 48 (2000) 1529e1538.
[7] (a) J.A. Raleigh, A.J. Franko, C.J. Koch, J.L. Born, Br. J. Cancer 51 (1985) 229e235;
(b) J.D. Chapman, A.J. Franko, J.A. Sharplin, Br. J. Cancer 45 (1982) 694e699;
(c) B.M. Garrecht, J.D. Chapman, Br. J. Radiol. 56 (1983) 745e753;
(d) J.S. Rasey, K.A. Krohn, Z. Grunbaum, P.J. Conroy, K. Bauer, R.M. Sutherland,
Radiat. Res. 102 (1985) 76e85;
(e) Z. Li, T. Chu, X. Liu, X. Wang, Nucl. Med. Biol. 32 (2005) 225e231;
(f) M. Dai, W. Zhu, Y. Xu, X. Qian, J. Fluoresc 18 (2008) 591e597;
(g) W. Zhu, M. Dai, Y. Xu, X. Qian, Bioorg. Med. Chem. 16 (2008) 3255e3260.
[8] A.C. Begg, E.L. Engelhardt, R.J. Hodgkiss, N.J. McNally, N.H.A. Terry,
P. Wardman, Br. J. Radiol. 56 (1983) 970e973.
[9] P.L. Olive, R.E. Durand, Cancer Res. 43 (1983) 3276e3280.
[10] P.L. Olive, Int. J. Radiat. Oncol. Biol. Phys. 10 (1984) 1357e1360.
[11] M.R.L. Stratford, E.D. Clarke, R.J. Hodgkiss, R.W. Middleton, P. Wardman, Int. J.
Radiat. Oncol. Biol. Phys. 10 (1984) 1353e1356.
[12] P. Wardman, E.D. Clarke, R.J. Hodgkiss, R.W. Middleton, J. Parrick,
M.R.L. Stratford, Int. J. Radiat. Oncol. Biol. Phys. 10 (1984) 1347e1351.
[13] A.C. Begg, R.J. Hodgkiss, N.J. McNally, R.W. Middleton, M.R.L. Stratford,
N.H.A. Terry, Br. J. Radiol. 58 (1985) 645e654.
4.2. Cell culture
Compounds were initially dissolved at 1 ꢂ 10ꢁ2 M in dimethyl
sulfoxide (DMSO), and small volumes were added to cell suspen-
sions to give the appropriate concentration. The final concentration
of DMSO was 1% or less.
V79 379A Chinese hamster cells were maintained as
exponentially-growing suspension cultures in Eagle’s Minimal
Essential Medium with Earle’s salts, modified for suspension
cultures with 7.5% fetal calf serum. The compounds were added to
cell suspensions to give the appropriate drug concentration. Then
the suspension was incubated in special gases (air þ 5% CO2,
nitrogen þ 5%CO2) at 37 ꢀC. After various periods of incubation with
these compounds, cell samples were removed centrifuged and
washed with PBS to remove residual compound, re-suspended in
a small volume of PBS, and evaluated by Fluorescence Microscopy
and Fluorescence Microplate Reader using appropriate excitation
wavelength.
[14] (a) P.L. Olive, D.J. Cha-din, Int. J. Radiat. Oncol. Biol. Phys. 12 (1986)
1247e1250.
[15] R.J. Hodgkiss, R.W. Middleton, M.R.L. Stratford, R.D. Buono, Biochem. Pharm-
col. 36 (1987) 1483e1487.
[16] M.F. Braña, M. Cacho, A. Gradillas, B. de Pascual-Teresa, A. Ramos, Curr. Pharm.
Des. 7 (2001) 1745e1780.
[17] (a) M. Hockel, K. Schlenger, C. Knoop, P. Vaupel, Cancer Res. 51 (1991)
6098e6102;
(b) W. Mueller-Klieser, P. Vaupel, R. Manz, R. Schmidseder, Int. J. Radiat. Oncol.
Biol. Phys. 7 (1981) 1397e1404;
(c) P. Vaupel, K. Schlenger, C. Knoop, M. Hockel, Cancer Res. 51 (1991)
3316e3322.
Acknowledgements
[18] (a) W.A. Denny, W.R. Wilson, M.P. Hay, Br. J. Cancer 74 (1996) 32e38;
(b) W.A. Denny, W.R. Wilson, J. Pharm, Pharmacology 50 (1988) 387e394.
[19] J.M. Brown, W.R. Wilson, Nat. Rev. Cancer 4 (2004) 437e447.
[20] (a) L.H. Patterson, Cancer Metastasis. Rev. 12 (1993) 119e134;
(b) A.K. Sinhababu, D.R. Thakker, Adv. Drug. Deliv. Rev. 19 (1996) 241e273.
[21] (a) K. Král, P. Zdenk, Microelectron. Engineer 69 (2003) 256e260;
(b) E. Gaiauskas, K.H. Feller, J. Lumin 102-103 (2003) 13e20;
(c) K.H. Feller, E. Gaiauskas, J. Lumin 83 (1999) 265e270;
(d) K.R. Naqvi, T.B. Melo, B.B. Raju, T. Javorfi, G. Garab, Spectrochim. Acta Part
A: Mol. Biomol. Spectrosc. 53 (1997) 1925e1936.
This work was financially supported by the Key New Drug
Creation and Manufacturing Program (2009ZX09103-102), the
China 111 Project (grant B07023) and the Shanghai Leading
Academic Discipline Project (B507) and the Shanghai Foundation of
Science and Technology (073919109).
References
[22] W. Chen, Y. Qin, Z. Cai, C. Chan, G. Luo, Z. Jiang, Bioorg. Med. Chem. 13 (2005)
1859e1866.
[23] (a) R.J. Hodgkiss, A.C. Begg, R.W. Middleton, J. Parrick, M.R.L. Stratford,
P. Wardman, G.D. Wilson, Biochem. Pharmacol. 41 (1991) 533e541;
(b) R.J. Hodgkiss, R.W. Middleton, J. Parrick, H.K. Rami, P. Wardman,
G.D. Wilson, J. Med, Chem 35 (1992) 1920e1926;
[1] R.S. Bush, R.D.T. Jenkin, W.E.C. Allt, F.A. Beale, H. Bean, A.J. Dembo, J.F. Prinde,
Br. J. Cancer. 37 (1978) 302e306.
[2] I. Churchill-Davidson, C.A. Foster, G. Wiernik, C.D. Collins, N.C. Piezez,
D.B.L. Skeggs, P.R. Purser, Br. J. Radiol. 39 (1966) 321e331.
[3] L.H. Gray, A.D. Conger, M. Ebert, S. Hornsey, O.C.A. Scott, Br. J. Radiol. 26 (1953)
638e648.
(c) R.J. Hodgkiss, J. Parrick, P. Manuchehr, M.R.L. Stratford, J. Med. Chem. 37
(1994) 4352e4356.
[24] W.R. Wilson, P. van Zijl, W.A. Denny, Int. J. Radiat. Oncol. Biol. Phys. 22 (1992)
693e696.
[25] D.C. Heimbrook, A.C. Sartorelli, Mol. Pharmacol. 26 (1986) 168e172.
[26] Y. Xu, B. Qu, X. Qian, Y. Li, Bioorg. Med. Chem. Lett. 15 (2005) 1139e1142.
[27] X. Qian, Y. Li, Y. Xu, Y. Liu, B. Qu, Bioorg. Med. Chem. Lett. 14 (2004)
2665e2668.
[4] R.J. Hodgkiss, Anti-Cancer Drug Des 13 (1998) 687e702.
[5] (a) J.A. Raleigh, M.W. Dewhirst, D.E. Thrall, Semin. Radiat. Oncol. 6 (1996) 37e45;
(b) M. Bache1, M. Kappler, H.M. Said, A. Staab, D. Vordermark, Curr. Med. Chem.
15 (2008) 322e338;
(c)K. Gulliksrud, I.K. Vestvik, K.Galappathi, B.Mathiesen, E.K. Rofstad,Radiat. Res.
170 (2008) 638e650;
[28] Y.S. Bcheshtiha, M.M. Heravi, F.A. Salehi, Indian. J. Chem. 37B (1998)
1177e1179.
(d) A. Cheguillaume, J. Gillart, D. Labar, V. Grégoire, J. Marchand-Brynaert, Bioorg.
Med. Chem. 13 (2005) 1357e1367.